Agent-based Simulation of UAV based Logistics Networks with Real World Data

Applications of Unmanned Aerial Vehicles (UAVs) are on the rise. Particularly within the healthcare sector the potential is huge as its cited as the most accepted application. This paper introduces an agent-based simulation to evaluate the network performance of UAV-based logistics networks in healthcare. The simulation is applied to a hypothetical real-world network. During a simulated day, the UAV fleet performs 212 flights, including 97 delivery flights, amounting to 4264 minutes enroute and covering a distance of 5941 kilometers. The analysis reveals average non-idle and mission utilization of 66% and 33%, respectively. The study also calculates annual network costs of EUR 2.23Mn, with a majority of it being direct costs (54.5%). Further sensitivity analysis identifies the biggest influences of battery capacity, C-Rate, and operator-to-UAV ratio on network performance and costs, highlighting these factors as critical for future optimization. Additionally, the benefit of incorporating various different UAV types into the network is only given if each UAV provides a unique value proposition to enhance the network performance.